• Havoc
  • Caius
  • redboot
  • Rules
  • Chain of Command
  • Members
  • Supported Ladders & Games
  • Downloads


Newbie Questions
09-19-2019, 06:24 AM,
#21
RE: Newbie Questions
(09-18-2019, 07:24 PM)Mowgli Wrote: Thanks a lot. As your "5 casualties on average" example demonstrates, the different methods of firing (if they're actually in use by the engine!) can lead to significant differences (for the chance to force a morale check on the target). In your example, letting units fire individually had a 7.6% greater chance (57.8%) to force a morale check on the target than letting the same units fire in a combined manner (50%). 

Then again, you'd probably also have to consider that if, e.g. 3 units fire individually, there are 3 randomization rolls/processes (random number between cv*low/high threshold), rather than just one? In other words, the result of a 3x "combined fire" would be more random than the result of 3x "individual fire".

As far as I can tell, there is only one Morale Check for an attack, whether by a single unit or a group. If this was not the case then an attack by a group could result in multiple checks and multiple failures. I do not recall seeing an attack that caused a unit to become Broken when it was not already Disrupted. But obviously you can achieve this by attacking multiple times with single units.

The only other explanation I can think of would be that there are multiple checks but the effects are not cumulative. In other words if there were multiple checks that gave one or more Disruption results then the the unit simply becomes Disrupted.

But the key thing, regardless how the engine does it calculations, is that firing individually results in additional chances to cause a Morale Check. When the casualties are higher than the example I gave of 5 men, then this becomes more pronounced.

The issue of whether the losses are more random when firing as a group does depend on how the engine does it calculations. A combined combat calculation would result in a more variable result, as you say, but they difference would not normally be great. Anyway, more variable is not necessarily bad. Depends if you are feeling lucky!
Quote this message in a reply
09-19-2019, 01:43 PM,
#22
RE: Newbie Questions
Hmm, so I had never seen that part of the manual before, that I recall. I have always worked from this part of the manual:

Disruption - if the fire result affects the target unit, then the target unit performs a Morale Check based on the amount of the effect.

So this states there is a morale check, but I can see the other section is different. Either way, I would guess it means the amount of losses compared to size impact the chance of failing the morale check, either by suppressing some results directly, or simply not making the check - the effect would be the same. Ultimately, a lot of the under the hood math in the game is not fully disclosed, and I have just learned to play.

But ultimately, I am fairly confident the fire losses are calculated separately, not summed and the results calculated as a whole. As the non-map results show each unit's fire separately, I am reasonably sure from prior discussions over the years that each one is calculated separately and the total losses are added up, rather than one calculation being applied.

Rick
[Image: exercise.png]
Quote this message in a reply
09-19-2019, 05:38 PM, (This post was last modified: 09-19-2019, 05:41 PM by Mowgli.)
#23
RE: Newbie Questions
Pls. ignore/delete - I have difficulties handling the quotes in this forum
Quote this message in a reply
09-19-2019, 05:42 PM, (This post was last modified: 09-19-2019, 06:38 PM by Mowgli.)
#24
RE: Newbie Questions
@ Green

You've never seen a unit go to broken by a single combined fire action. That's a good point as it (as far as I understand it) indicates quite clearly that in a "combined units" fire action, the fire power of several units is summed up for a single action rather than several small actions.

I'm currently working on a little guide/manual in German for Panzer Campaigns, here is a quick translation of the "morale check" part:


Quote:If a unit has suffered at least one casualty as the result of a combat action, it needs to test whether it is forced to undergo a morale test. The more losses the unit had to take and the smaller the unit, the more likely it is that the unit is forced to make a morale check. Here is a table to give you an impression:  

suffered losses* = chance to force morale check
small unit (1 component) 5=50%, 10=66%, 15=75%, 30=86%
medium unit (2 components) 5=33%, 10=50%, 15=66%, 30=75%
large unit (3 components) 5=25%, 10=40%, 15=50%, 30=66%
*The loss of a vehicle or gun counts as 10

Some weapons/units or situations are more likely to force a morale check on some units (even if the inflicted losses are comparatively small)**:  
• HQ-units and hard targets (armored units, deployed infantry in bunkers or pillboxes) are vulnerable against artillery fire.  
• Tank units are vulnerable against non-self propelled anti tank units and non-self propelled heavy anti air units.
• Units who attack (not defend) in an assault are vulnerable. Especially at night [WIP: difficult terrain mitigates this malus somehow...]
** If a unit is "vulnerable", then the losses it has taken are doubled to determine the chance whether it needs to perform a morale check.

IF a unit is forced to perform a morale check AND IF the unit fails the morale check, it becomes disrupted or - if it's currently already disrupted and is at maximum fatigue - broken.




COMBAT VALUE IN FIRE COMBAT

I've taken a closer look at the numbers in the very usefull "results window". Note I'm only refering to infantry engagements for now.
  • "Range" obvisously refers to the range.
  • "Modifier" indiciates the terrain bonus of the target.
  • "Mod" refers to the quality modifier (of the firing unit) that is applied to the cv.
  • "inf x%" indicates the strength of the unit (it's always higher than the actual strength, see manual, p. 38 --> infantry fire effectiveness)
I've also figured out that the the given number does already take into account the malus on defense that units get for being in travel mode, which also tells me that the given number already includes the (/defensive stat) part of the formula.

The problem I have, however, is that I still can't re-create the combat value which is shown in the window. This wouldn't be a big problem if my own estimations were somewhat close to the actual CV. But unfortunately, that's not always the case. My calculations are WAY off. Note that the cv shown in the game are very consistent, so there is no random factor. 

Quote:Perhaps an example can help to demonstrate my problem:
An infantry unit of 75 men (nominal 88%, so that means the max strength would be 85), quality F, soft fire 6, no fatigue, no disruption or other conditions
fires at
a soft target unit with defense 16, in travel mode (terrain is ignored), no hexside obstacles, no elevation difference

The actual result looks like this:
range = 1, Modifier = 0%
Fire = 60 (mod = -60%, inf = 95%)

whereas my own calculations suggest a very different result:
85 /10 *0.95 *6 *0.4 /16 *2
max.strength / 10 (infantry) * effective strength (as clearly indicated by the "inf" information) * soft fire * quality F / defense of target * 2 (target in travel mode)
= 2.4

Now obviously, there is a big difference between 60 and 2.4. If, for some reason, you don't divide by the defense of the target, then the result gets closer, but is still significantly off (38.4). 

Crazy

On the other end of the extreme, I once had a tank unit fire at infantry and my own understanding of the formula promised me a cv of over 300 (without /defense) or ca. 19 (with /defense) . The actual cv was 150. As 150 is a suspicious value for a unit of 15 tanks, maybe there is an upper cap to the cv (cv capped to 10 * max. strength?)? The same could be true for the case above. The combat value changed to 30 when the target was not in travel mode (terrain was clear). 30 is 15*2. So maybe there is also a low cap for fire combat?

I don't want to crunch too many numbers. But the difference between what I expect (according to the rules) and what I get are simply too big here. I want to develop a base feeling for the game.
Quote this message in a reply
09-19-2019, 09:45 PM, (This post was last modified: 09-20-2019, 07:36 AM by Green.)
#25
RE: Newbie Questions
(09-19-2019, 05:42 PM)Mowgli Wrote: The problem I have, however, is that I still can't re-create the combat value which is shown in the window. This wouldn't be a big problem if my own estimations were somewhat close to the actual CV. But unfortunately, that's not always the case. My calculations are WAY off. Note that the cv shown in the game are very consistent, so there is no random factor.

I understand your desire to fully comprehend the rules but my advice would be to take a step back from the detail.

The Fire Report is not a perfect window into the combat calculation. It is a long time since I looked at it but what exactly is included in the CV is not always clear. The first thing to note is that you do not divide the number of men by 10 but instead multiply the number of vehicles by 10. Not what the manual says but what I believe you need to do or your CV will be out by a factor of 10. Secondly I think the CV shown is before the modifiers given in the brackets. But I am going from memory and may have this wrong. All I remember for sure is that it was all quite complex.  

Anyway, making these two changes to your calculations would result in a revised figure of 63.75. Still not right but in the ball park. But perhaps this is a coincidence as I wonder if the defense value of 16 is correct? This normally drops to something like 10 or 12 when changing to Travel mode if the unit has transport of some sort.

The reality is that there is a lot going on and trying to replicate the combat calculation is a difficult and probably pointless task. When you play the game you just need to understand the concepts.

Edit: My comment above that "the CV shown is before the modifiers given in the brackets" is only partially correct. Actually the 'inf %' shown in the brackets is applied to the CV. So for your calculation the CV shown on the Fire Report is determined as;

 85*.95*6/16*2 = 60.56

But is this useful? Most combat is much more complicated than this and even this figure does not include the Quality modifier. Few players ever look at anything other than the on-Map results, and for good reason. This level of detail beyond what you need to know to make sensible decisions and is more likely to be a distraction.
Quote this message in a reply
09-19-2019, 10:12 PM,
#26
RE: Newbie Questions
(09-19-2019, 01:43 PM)Ricky B Wrote: Hmm, so I had never seen that part of the manual before, that I recall. I have always worked from this part of the manual:

Disruption - if the fire result affects the target unit, then the target unit performs a Morale Check based on the amount of the effect.

So this states there is a morale check, but I can see the other section is different. Either way, I would guess it means the amount of losses compared to size impact the chance of failing the morale check, either by suppressing some results directly, or simply not making the check - the effect would be the same. Ultimately, a lot of the under the hood math in the game is not fully disclosed, and I have just learned to play.

But ultimately, I am fairly confident the fire losses are calculated separately, not summed and the results calculated as a whole. As the non-map results show each unit's fire separately, I am reasonably sure from prior discussions over the years that each one is calculated separately and the total losses are added up, rather than one calculation being applied.

Rick

Sorry Rick, but we seem to be talking at cross purposes.

The manual states;

"Thus a battalion unit that takes a loss of 15 men has a 50% chance of requiring a morale check and a battalion unit that takes a loss of 60 men has about an 80% chance of requiring a morale check."

Are you saying the manual is in error? This cannot be interpreted as meaning that the amount of losses impacts the chance of failing the Morale Check. It is unambiguous that the losses impact the chance of requiring a Morale Check.

I understand that the manual says;

"Disruption - if the fire result affects the target unit, then the target unit performs a Morale Check based on the amount of the effect."

But I think this is just poorly worded and should be interpreted as;

"Disruption - if the fire result affects the target unit, then based on the amount of the effect, the target unit performs a Morale Check."

The manual would then be consistent.

Whether the losses are calculated separately or together is not the issue as the average losses would be the same either way. It is not surprising therefore that "prior discussions over the years" have found no fault with this.

What is at issue is how many Morale Checks are performed. My point is that there is never more than one Morale Check performed per attack. I have never seen an attack cause a non-Disrupted unit to go straight to being Broken. So I do not see how "the effect would be the same". Clearly, attacking as a group or attacking separately can yield very different results in terms of Morale Checks. Are you saying you do not agree with this?
Quote this message in a reply
09-19-2019, 11:31 PM,
#27
RE: Newbie Questions
Regarding your question about a game with maneuver and a smaller unit count-El Alamein is a good fit.
Quote this message in a reply
09-20-2019, 01:52 PM,
#28
RE: Newbie Questions
(09-19-2019, 10:12 PM)Green Wrote: [quote pid='432494' dateline='1568864614']

What is at issue is how many Morale Checks are performed. My point is that there is never more than one Morale Check performed per attack. I have never seen an attack cause a non-Disrupted unit to go straight to being Broken. So I do not see how "the effect would be the same". Clearly, attacking as a group or attacking separately can yield very different results in terms of Morale Checks. Are you saying you do not agree with this?

[/quote]

So my last comments were clearly about losses and not morale checks. However, I also believe that each one goes through the morale check process separately. However, they also are calculated based on the initial pre-fire status, as if all happen at the same time, from what I understand (having worked with the early designers and all). I don't even know how to test it with absolute certainty. However, remember to break a unit has to be at max fatigue. So it could be possible to have a test where a unit is at say 299 fatigue but not disrupted and fired at by separate strong units in a combined fire, each of which cause heavy losses (so maybe modified combat values on both sides), and see if it ever breaks.

But if the idea is to try and determine the best way to maximize things, just remember (and you can find lots of discussion here already on it) that there are trade off between firing all units in a stack together versus one at a time (which definitely could result in a disruption then broken status change in one series of shots). Ultimately better play would overwhelm any gain from perfect understanding of the math involved, and if it is purely curiosity, again I don't believe there is a way to prove it without the code. I know even the basic CV calc is much more complex than the manual states, and there are so many factors that impact it that we haven't been able to get a grip on them all in one shot.

Rick
[Image: exercise.png]
Quote this message in a reply
09-24-2019, 02:13 AM, (This post was last modified: 09-24-2019, 04:10 AM by Mowgli.)
#29
RE: Newbie Questions
Small update:

I think the missing link in my calculations of fire combat is related to unit size. I haven't had the time to check it systematically yet, but I might be on to something (there is a system in the deviations of my calculations to the numbers shown in the game).

My suspicion is that there is a factor that gets added (in addition to the ordinary strength factor) based on unit size. I don't think that this is actually related to the "unit category" (pltn/coy/btn). I think it's just related to (on paper) strength of the component. 

Examples: 

I had a large unit (soviet rifle btn, component strength on paper 450 men) fire at a german motorcycle company. The resulting fire value was *1.4 of what I was expecting (if you calculate the result according to what the manual says). Note that this did not change when I combined several motorcycle companies into a combined unit (+++). The result was also multiplied by 1.4, no matter whether the target consisted of a single or 2 or 3 combined motorycle coys.

The motorcycle coys firing at the rifle btn., however, fired at *1, that is: as my calculations and the manual suggested. (Again, it did not matter how many of them were combined into a single unit).

Then I took a closer look at the same 450-rifle-btn firing at a german btn whose on paper component strength was  250 men. The result was *1.25. So the large rifle btn did not get as big a bonus as against the smaller companies.  

The german 250 man btn firing at the large russian rifle btn, however, also got a small bonus (*1.2).

Things get messed up when it comes to tanks. E.g. 13 tanks firing at a large infantry unit (450 on paper) also had their fire value multiplied by 1.4.

--------------------------------------
By the way this is how I calculate the fire values for infantry:

soft fire value
*max. paper strength
* effective strength % (as indicated in the combat report, not by the unit info)
* quality
* all other potential factors, including terrain and defenses of the target, disruption, target in travel mode, etc.
/ defense of target (watch out for changes due to travel mode if target is mech/motor. inf)

@ max paper strength: you need to look it up in the database, or use this method (pick the closest plausible nr.): "nr. of current men" / "current %" *100

In the instances mentioned above, these calculations seem to be quite precise (within ca. 5% of the actual fire result; the difference might be caused by differences in rounding up/down?)

--------------------------------------
So my hypothesis is that - at least for infantry? - larger units (classified by max. component strength?) get an additional bonus on their fire value. The larger their target, the smaller the bonus. It seems as if this bonus can be as big as *1.4. I suppose that tanks are classified differently (in general higher than the infantry?). I've not taken a look at arty or tanks vs tanks yet.

I will update this post once I've had time to take a closer look and have gained more insights.
Quote this message in a reply
09-24-2019, 07:53 AM,
#30
RE: Newbie Questions
(09-19-2019, 05:38 PM)Mowgli Wrote: Pls. ignore/delete - I have difficulties handling the quotes in this forum

Gent:  Smoke7

Maybe it's time to put your scratch pad away, join the TOC ladder and get some PBeM games underway?  Whistle
Regards, Mike / "A good plan violently executed now is better than a perfect plan executed next week." - George S. Patton /
Send this user an email
Quote this message in a reply


Forum Jump:


Users browsing this thread: 1 Guest(s)